pce-fws 20 problem

[cabillo]

First of all sorry for my english.
I have a pce-fws 20 (wh-1080 compatible) station that works on cumulus 1.8.9.
This is the cumulus site: http://www.palmimeteo.tk (it is located in Palmi, Calabria, Italy)
As you can see I have some error on the temperature graph, I thought it was a signal problem (the station is wireless, distance:20m, but there is a roof in the middle) but the humidity graph is ok, so I really don't know.
what's the problem?

[steve]

These stations do seem to produce temperature 'glitches' occasionally, and I don't think anyone knows why. There is a setting in the cumulus.ini file which you can use to try to get around these glitches. With Cumulus stopped, edit cumulus.ini, and in the [Station] section add a line like this:

EWTempdiff=x.y

where x.y is maximum difference allowed in temperature between each reading.

[cabillo]

These stations do seem to produce temperature 'glitches' occasionally, and I don't think anyone knows why. There is a setting in the cumulus.ini file which you can use to try to get around these glitches. With Cumulus stopped, edit cumulus.ini, and in the [Station] section add a line like this:

EWTempdiff=x.y

where x.y is maximum difference allowed in temperature between each reading.

Ok, I will try, thank you.

[cabillo]

cumulus.ini:

Code: Select all

[Station]
EWTempdiff=0.3 
Type=5
Model=wh-1080
Port=0
Latitude=38,3594444444444
Longitude=15,8480555555556
LatTxt=N 38° 21' 34"
LonTxt=E 15° 50' 53"
Altitude=230
AltitudeInFeet=0
Humidity98Fix=0
Wind10MinAverage=0
CalculatedDP=0
CalculatedWC=0
RolloverHour=0
Use10amInSummer=1
ConfirmClose=0
CloseOnSuspend=1
RestartIfUnplugged=0
SyncDavisClock=0
DataLogInterval=1
WindUnit=2
PressureUnit=1
RainUnit=0
TempUnit=0
LocName=Palmi (RC)
LocDesc=
StartDate=03/02/2010
WindComp=1
LastLoggerDownload=03/02/2009 21.30.46
YTDrain=0
YTDrainyear=0
EWInterval=0
EWFile=C:\Program Files\EasyWeather\EasyWeather.dat
UseDataLogger=1
UseCumulusForecast=1
UseCumulusPresstrendstr=0
FCpressinMB=1
FClowpress=950
FChighpress=1050
ForceVPBarUpdate=0
UseZeroBearing=0
VP2ConnectionType=1
VP2TCPPort=22222
VP2IPAddr=0.0.0.0
WarnMultiple=0
CloudBaseInFeet=0
UseSpeedForAvgCalc=0
[Web Site]
...

I tried with 0.3, but I got this:
on the right: 14.3° -> 13.2° -> 14.3°
What's wrong? I don't understand.
(I edited cumulus.ini with cumuls stopped)

[steve]

Your temperature reading didn't suddenly change from 14.3 to 13.2, it changed gradually over a period of several minutes. If you think that's not correct, then I think it's a problem with your sensor.

[cabillo]

Your temperature reading didn't suddenly change from 14.3 to 13.2, it changed gradually over a period of several minutes. If you think that's not correct, then I think it's a problem with your sensor.

I'm not sure of this, sensor update is very slow. (1minute)
I tried with 0.1, and looking steadily to the cumulus gui i have seen these are not gradual changes but instant changes of 0.3° and 0.2°:

[Gina]

These stations do seem to produce temperature 'glitches' occasionally, and I don't think anyone knows why.

Mine's suddenly got a lot worse and I'm investigating. I thought it was signal loss but I'm changing my mind. Interference is another suggestion (as discussed in another thread).

[steve]

Sorry, I hadn't noticed that you've used a decimal point in the cumulus.ini file, but your system is configured to use a comma, so you need to use a comma for the decimal for EWtempdiff.

[steve]

Mine's suddenly got a lot worse and I'm investigating. I thought it was signal loss but I'm changing my mind. Interference is another suggestion (as discussed in another thread).

Yes, I've been reading your reports with interest. Hopefully you'll find out what's causing it!

[Gina]

Yes, I hope so

[cabillo]

Yes, now it works with comma, thank you again, but this is what I got this morning.
This is not a lost sensor contact, the other things were updating, and the temperature on the hardware screen was updating.
I used EWtempdiff=0,3 , I think that if temperature changes more than 0.3° Cumulus keeps the previous temperature unchanged, so at 07.00 temperature changed more than 0.3° and continued to rise and Cumulus kept 13.0° (at 9.40 I restarted Cumulus) Am I wrong?
Is there a way that if temperature changes more than 0.3°, for example +0.6° Cumulus makes a +0.3° change always, and for -0.8° it makes a -0.3°?

[steve]

The EWtempdiff facility is quite basic and you have to be careful in your choice of setting. 0.3C is probably too low as it will lead to false triggers, and then, as you have seen, you are 'stuck' until the temperature changes sufficiently between readings to return to close to where it was before. Typically the 'real' spikes that these stations produce are of the order of several degrees, and many people are using this facility successfully to remove these spikes, with a reasonable choice of value.

[Carsten]

First of all sorry for my english.
I have a pce-fws 20 (wh-1080 compatible) station that works on cumulus 1.8.9.
This is the cumulus site: http://www.palmimeteo.tk (it is located in Palmi, Calabria, Italy)
As you can see I have some error on the temperature graph, I thought it was a signal problem (the station is wireless, distance:20m, but there is a roof in the middle) but the humidity graph is ok, so I really don't know.
what's the problem?

hi cabillo,
do you speak german ??

[cabillo]

First of all sorry for my english.
I have a pce-fws 20 (wh-1080 compatible) station that works on cumulus 1.8.9.
This is the cumulus site: http://www.palmimeteo.tk (it is located in Palmi, Calabria, Italy)
As you can see I have some error on the temperature graph, I thought it was a signal problem (the station is wireless, distance:20m, but there is a roof in the middle) but the humidity graph is ok, so I really don't know.
what's the problem?

hi cabillo,
do you speak german ??

No, I don't.